基于PC机的医学图像三维表面重建

目的：利用CT，MR等获得的图像，重构出器官、骨骼或组织的三维形体，帮助医生诊断、治疗或制订手术方案。方法：Shear-Warp算法通过将三维数据场的投影分解为错切变换（shear)和变形变换（warp)两步来实现三维重建，显著减少了计算量，先对体数据进行错切变换，然后在错切空间根据阈值法获取三维表面，根据光照模型得到三维表面的亮度，最后通过变形变换得到最终的结果图像，结果：提出了一种基于Shear-Warp投影原理快速显示三维表面的直接重建算法。结论：与其他表面重建算法比较，这种算法不需要预处理，重建过程中不必生成中间数据，在无硬件加速的条件下，能在普通微机上实现医学图像的快速三维表面重建。

Fast 3D Surface Rendering for CT or MR Image on a Personal Computer

Objective To design a 3D surface reconstruction algorithm capable of reconstructing 3D shapes of the organs, bones and tissues rapidly from CT or MR image on a personal computer. Method By transforming and decomposing the projection of the discrete 3D data field into shear transformation and warp transformation, a Shear Warp algorithm realizing the 3D reconstruction was established. The algorithm reduce the amount of computation obviously.The 3D surface was reconstructed by using the principle of the parallel projection,basing on the Shear Warp algorithm.Shear transformation was performed on the volume data at first. Then the warped 3D surface was obtained by making a threshold in the sheared space. The brightness of the 3D surface was calculated using Phong brightness model, and the final image was generated through the Warp transformation. Result A direct 3D surface reconstruction algorithm was established basing on a principle of projection from the Shear Warp algorithm, which was able to display 3D surface on a personal computer rapidly. Conclusion Compared with other surface reconstruction algorithms, the proposed algorithm has the advantages of no preprocessing, no intermediate data are generated in the reconstruction. With this new algorithm, it is possible to realize fast 3D surface rendering for medical images on a personal computer without the support of any hardware accelerator.